What are the requirements to use MCP Browser?
You need Node.js 18+ and either Chrome or Edge browser installed on your machine.
@cherchyk
Basic information
Browser automation for web scraping when standard HTTP requests fail. Use ONLY when pages require: (1) Authentication - login forms, SSO, 401/403 errors, corporate intranets, (2) Anti-bot protection - CAPTCHA, Cloudflare challenges, rate limiting, (3) JavaScript rendering - SPAs,
Add this server to your MCP-compatible client using the configuration below.
{
"mcpServers": {
"MCPBrowser": {
"type": "stdio",
"command": "npx",
"args": [
"-y",
"mcpbrowser@latest"
],
"description": "Browser automation for web scraping when standard HTTP requests fail. Use ONLY when pages require: (1) Authentication - login forms, SSO, 401/403 errors, corporate intranets, (2) Anti-bot protection - CAPTCHA, Cloudflare challenges, rate limiting, (3) JavaScript rendering - SPAs, dynamic content loaded after page load. Can fetch pages, click elements, fill forms, and extract content. Opens real browser for user authentication, then automates interactions and extraction. DO NOT use for simple HTML pages that work with regular HTTP GET requests."
}
}
}Tool list not available yet
This listing does not currently have a tool list to display. This does not mean the server has no tools; check its documentation for supported capabilities.
Fetch the live tool list by running this server in a temporary sandbox using the button above.
MCP Browser is an MCP (Model Context Protocol) server that gives AI assistants (Claude, Copilot) the ability to browse web pages using your real Chrome or Edge browser session. It handles authentication, CAPTCHAs, anti-bot protection, and JavaScript-heavy sites by leveraging your existing logged-in session rather than using headless automation.
Install via the VS Code Marketplace extension (one‑click), a single Claude Code command (claude mcp add mcpbrowser), manual configuration in Claude Desktop’s claude_desktop_config.json, or the npx -y mcpbrowser@latest command. After setup, the AI can invoke tools like fetch_webpage, click_element, type_text, get_current_html, and close_tab to interact with web pages.
removeUnnecessaryHTML is enabled.You need Node.js 18+ and either Chrome or Edge browser installed on your machine.
It launches your local Chrome or Edge in a managed mode and reuses your existing user data directory, so cookies and authentication persist across sessions.
Any MCP‑compatible host, including Claude Desktop, Claude Code (CLI), and GitHub Copilot. The VS Code extension auto‑configures it for Copilot.
Yes. Load the login page with `fetch_webpage`, then use `type_text` and `click_element` to log in. The session stays open, and you can call `get_current_html` or further tools on the same domain without re‑authenticating.
Use `close_tab` with the page’s URL. This closes the associated browser tab and forces a fresh session the next time that domain is visited.

Parse allows your agent to access thousands of APIs instantly. You can use Parse to build an API for any action on any website and edit any previously existing API in our catalogue.

Model Context Protocol Servers
Capture screenshots of webpages as images or PDFs with Screenshot Scout.
Give your AI agent the live web. One MCP server for Search, Fetch, Browser, and Agent: browse real sites, extract structured data, fill forms, and run multi-step workflows. Search and Fetch are free.
A flexible HTTP fetching Model Context Protocol server.
Comments